You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
ClinicalTrialsDataProcessing/market_data/extract_nsde.py

86 lines
2.3 KiB
Python

import json
import psycopg2 as psyco
from psycopg2.extras import execute_values
import datetime as dt
def file_generator(max_num):
for itt in range(1,max_num+1):
filename = "other-nsde-{:0>4}-of-{:0>4}.json".format(itt,max_num)
yield filename
def get_date(result,key):
r = result.get(key)
if r:
return dt.datetime.strptime(r, "%Y%m%d")
else:
return None
def build_values(result):
#adjust types
proprietary_name = result.get("proprietary_name")
application_number_or_citation = result.get("application_number_or_citation")
product_type = result.get("product_type")
package_ndc = result.get("package_ndc")
marketing_category = result.get("marketing_category")
package_ndc11 = result.get("package_ndc11")
dosage_form = result.get("dosage_form")
billing_unit = result.get("billing_unit")
marketing_start_date = get_date(result,"marketing_start_date")
marketing_end_date = get_date(result, "marketing_end_date")
inactivation_date = get_date(result, "inactivation_date")
reactivation_date = get_date(result,"reactivation_date")
return (
proprietary_name
,application_number_or_citation
,product_type
,package_ndc
,marketing_category
,package_ndc11
,dosage_form
,billing_unit
,marketing_start_date
,marketing_end_date
,inactivation_date
,reactivation_date
)
if __name__ == "__main__":
for x in file_generator(3):
#It would be nice to replace this^^ file_generator with something that retrieves and unzips the files directly.
with (
psyco.connect(dbname="aact_db", user="root", host="localhost", password="root") as con,
con.cursor() as curse,
open("./"+x,"r") as j
):
print(x)
results = json.loads(j.read())["results"]
query = """
INSERT INTO spl.nsde (
proprietary_name
,application_number_or_citation
,product_type
,package_ndc
,marketing_category
,package_ndc11
,dosage_form
,billing_unit
,marketing_end_date
,marketing_start_date
,inactivation_date
,reactivation_date
)
VALUES %s;
"""
values = [build_values(y) for y in results]
execute_values(curse,query,values)